← Back

Polling tracker

What follows if support or turnout shifts: change any party’s share, put parties on joint lists, or move turnout and support in any group of voters (men aged 18–29 in towns, say), and see the seats and the options for government, with the model’s uncertainty. Every scenario has a link of its own.

How this works

Parties and lists: when you edit a vote share or build a joint list, your shares are applied to each of the model’s simulated draws for that week (each party’s draws are scaled so that their median is the share you entered), the voters’ response to joint lists (below) is applied draw by draw, and the seat simulation is run again in your browser. A joint list’s total has a slider of its own: moving it scales the list’s parties in proportion, and moving one of its parties resizes the others so that the list keeps its total. Scenarios therefore keep seat ranges and probabilities: they describe what would follow if support stood where you set it, with the same uncertainty as the model. Government options follow the compatibility rules described on Seats with joint lists, two lists can govern together only if no party on one is excluded from a party on the other.

Voter response to joint lists: when parties stand on one list, some of their voters will not follow. For each party on a joint list, the tracker takes the share of its voters who would rate each partner in the bottom quarter of a 0 to 10 scale, from a model that combines surveys of 2023 to 2026: the Polish General Election Study (2023 and 2025), the European Election Study (2024), CBOS and SWPS University’s online surveys. Each survey is read through its own scale, the 2023 Polish General Election Study’s ratings of the parties set the scale, and feelings across the left-right divide can move over time with CBOS’s monthly trust in politicians. A share of those reluctant voters defect: 40% in the Medium setting, calibrated on CBOS’s questions about joint opposition lists in February 2023, with 30% and 55% (Low and High) at the 10th and 90th percentiles of that estimate. Defectors abstain or vote for the parties outside the list that their electorate would most consider (from the European Election Study 2024). The reluctance of each pair and the abstention share vary from draw to draw, so the scenario keeps its uncertainty. R+, which no survey rated, takes the figures of PiS. The abstention share and the extra reluctance between parties that split recently (PiS and R+, Polska 2050 and PSL) are judgement, not estimates, and the advanced settings let you change them.

Groups: the 2023 electorate is 280 cells of age, sex, education and place of residence, from the Polish General Election Study of 2023 (1299 respondents, through the vote model of my Voting behaviour analysis). Each cell has a chance of voting and a vote among the 2023 lists, raked to PKW’s turnout and list shares and to the pattern across groups in Ipsos’s exit poll for TVP, TVN and Polsat, which is first moved onto the official result (the table below). The starting point is this week’s estimates, with each group moved as Opinia24′s and CBOS’s tables show it has moved since December 2023 (or, if you choose, every group moved alike), each splinter taking a share of its 2023 list that differs by group as those tables show, and each group’s turnout moved as its stated intention to vote has moved, with national turnout kept at 2023′s. A change sets turnout, or one party’s share of the voters, in a group or an intersection of groups; the cells are raked in log-odds until every change is met, everything not set moving as little as possible, and the national result follows from the groups’ sizes. Those national shares become the scenario’s shares, and the seats follow as for any scenario. Parties set by hand under Parties and lists stay where they were set. The regional patterns do not yet follow the groups, so a change among the young moves every constituency alike. Changes by group start from the week the groups were last fitted to the polls.

Who votes: for every group, everyone in it votes (turnout raised to 100%) or only it votes (everyone else stays at home), from the same starting point; the extra voters are assumed to vote like the voters already in their cell, so the figures show what composition alone can do. Seats there are a single allocation by the tracker’s rules, without joint lists.

The figures describe what would follow if support stood where you set it, not a forecast. How the polls are pooled is on Polls.

The 2023 targets: Ipsos’s exit poll, moved onto the official result

Each row is one figure the cells are raked to: the exit poll as printed, the target after moving the exit poll’s pattern onto PKW’s result, and the survey model before raking. Turnout by residence comes from the late exit poll (90% of stations); everything else from the 21:00 exit poll. Each figure links to the report it was read from.

MeasureGroupListExit pollTargetSurvey modelSource
TurnoutMen7273.576.9report
TurnoutWomen73.775.272.3report
Turnout18-2968.870.363.3report
Turnout30-3972.373.773.5report
Turnout40-4978.579.774.9report
Turnout50-5983.284.278report
Turnout60 and over6768.677.6report
TurnoutVillage70.370.775.6late poll
TurnoutTown up to 50,00074.174.573.4late poll
TurnoutCity 51,000-200,00073.974.374.2late poll
TurnoutCity 201,000-500,00082.682.974.2late poll
TurnoutCity over 500,00081.281.574late poll
VoteWomenPiS36.535.336.7report
VoteWomenKO3332.131.7report
VoteWomenTrzecia Droga1314.515report
VoteWomenLewica10.110.29.5report
VoteWomenKonfederacja3.443.5report
VoteMenPiS37.135.533.7report
VoteMenKO30.229.130report
VoteMenTrzecia Droga12.914.313.8report
VoteMenLewica6.96.97.7report
VoteMenKonfederacja9.310.810.9report
Vote18-29PiS14.913.112.5report
Vote18-29KO28.325.832.5report
Vote18-29Trzecia Droga16.918.220.4report
Vote18-29Lewica17.717.612.1report
Vote18-29Konfederacja16.619.917.5report
Vote30-39PiS26.323.815.5report
Vote30-39KO29.527.635.6report
Vote30-39Trzecia Droga17.118.921.5report
Vote30-39Lewica10.610.810.7report
Vote30-39Konfederacja1113.511.4report
Vote40-49PiS3229.722.1report
Vote40-49KO35.333.937.8report
Vote40-49Trzecia Droga15.317.317.6report
Vote40-49Lewica8.18.59.4report
Vote40-49Konfederacja4.86.19report
Vote50-59PiS44.241.835.2report
Vote50-59KO32.932.235.6report
Vote50-59Trzecia Droga11.813.612.9report
Vote50-59Lewica5.15.59report
Vote50-59Konfederacja2.93.74.3report
Vote60 and overPiS535159.8report
Vote60 and overKO31.331.221.7report
Vote60 and overTrzecia Droga7.58.87.8report
Vote60 and overLewica5.25.75.8report
Vote60 and overKonfederacja11.32.1report
VotePrimary or lower secondaryPiS63.756.668.1report
VotePrimary or lower secondaryKO15.617.912.6report
VotePrimary or lower secondaryTrzecia Droga6.99.17.3report
VotePrimary or lower secondaryLewica4.55.43.5report
VotePrimary or lower secondaryKonfederacja5.46.94.6report
VoteBasic vocationalPiS62.255.146.9report
VoteBasic vocationalKO18.220.823.6report
VoteBasic vocationalTrzecia Droga7.61013.2report
VoteBasic vocationalLewica4.257report
VoteBasic vocationalKonfederacja4.25.46.6report
VoteSecondaryPiS38.531.831.9report
VoteSecondaryKO30.432.430.8report
VoteSecondaryTrzecia Droga1214.716.3report
VoteSecondaryLewica8.199.7report
VoteSecondaryKonfederacja78.38.1report
VoteHigherPiS22.617.818.1report
VoteHigherKO39.640.343.3report
VoteHigherTrzecia Droga16.519.315.6report
VoteHigherLewica11.211.910.4report
VoteHigherKonfederacja6.37.27.4report
VoteVillagePiS48.646.740.8report
VoteTown up to 50,000PiS34.33333.2report
VoteCity 51,000-200,000KO37.736.636.1report
VoteCity 201,000-500,000KO42.641.432.9report

Ben Stanley, 2026. Polling tracker, Polish Political Analyses, benstanley.eu.